MyTfL: what we have learned

The cover story

Yesterday morning a London travel planner app appeared in the iTunes app store called MyTfL. Nothing unusual there, there are already half a dozen available. However, MyTfL was different: it described itself as the ‘official’ Transport for London app, and it appeared to contain data feeds (e.g. live bus departures) that are not currently available to developers. This was evidently a significant development: so why was this it so quietly released with no fanfare?

On further examination the mystery deepened. The publisher was actually Mentz Datenverarbeitung Gmbh, who are a German system integrator working for TfL Group Marketing, who run the TfL Journey Planner. And the app used the full TfL branding, which is strictly controlled (see here for details). So what was this new app and how did it come to be released with hitherto unreleased data?

The implications

The short answer is that we don’t know what is going on here on the outside. Or even on the outside of the inside for those of us on the Mayor’s Digital Advisory Board. However, we can speculate on our slender knowledge. The key guess must be that all data that developers would like to access to make new apps is already polished and ready inside TfL. When we are told that it is not available for release then we must take this with a giant pinch of salt. Another guess must be that this is the equivalent of the ‘bureaucrats strike back’: the thinking presumably being that one way to stymie these pesky developers is to shoot their fox. With all the data in a pile, lots of public money and a few iPhone developers it must be easy to render independent attempts to create value from the datasets unnecessary and futile.

Finally, I would also guess that MDV were given this opportunity because they are the ones with most to lose from an open market in apps leveraging open transport data and so they were keen to take on the challenge. If developers were given access to the TfL Journey Planner API (search the web for the jailbreak access) then MDV would be enabling other commercial players to make money from their own work. However, the whole point of opendata access to government data is to break down stifling monopoly access to data to allow innovation around the data. If TfL release an all-singing app then it is hard for developers to justify the investments needed to do that innovation because the state has intervened in the market with taxation funding. With opendata access to all the live data feeds there is no market failure and no case for intervention by the state. Both this government and the last accepted that there is more value to UK PLC by allowing innovation (leading to increased tax receipts) than by allowing government departments to trade data to recover the cost. So this kind of app release is against all current policy.

Evidence, if we needed it, that this was a controversial release, came when equally mysteriously, the app disappeared from the iTunes app store today. From which we are forced to conclude that this was a freelance operation not cleared with TfL senior management. So there you have it: a brief glimpse within the Data Wars kimono, and a microcosm of political and economic conflicts to come around apps and data.

The backstory

The app itself is (was?) a conventional journey planner app in which you enter your origin and destination, the journey time and the modes to search:

IMG_0021.PNG IMG_0020.PNG

The app also carried live departures for every mode of London Transport: tube, DLR, tramlink, bus and river. It also carries live departures from National Rail… yet there are no acknowledgments to National Rail Enquiries. It is the bus live departures that are of greatest interest as they are not available anywhere except central London bus stops on Countdown. But here they are in an app:


There are omissions… there is nothing about the Cycle hire scheme. But the app does do cycle routes, though not nearly as well as CycleStreets: they are drawn 1 pixel wide in yellow on a beige background and doesn’t avoid main roads very well. Here are the routing result page and the ‘interactive’ map page:

IMG_0025.PNG IMG_0024.PNG

So this is not the app to set the world on fire, and there are alternative like Journey Pro from Navitime. It is more important for what it reveals about the battle to convert government bureaucracies to the cause of opendata, even if the political direction is clear.

Jonathan Raper

5 thoughts on “MyTfL: what we have learned

  1. Steven Feldman

    Sounds like an FoI request is needed here.

    Questions might include – did TfL pay MDV to build this app? If so what procurement process? Why just iPhone? If not paid under what conditions?

    That’s for a start. I guess @charlesarthur would know how to frame the questions.

    Ratty smell here?


  2. Mark

    I also get the impression that there’s serious inertia (if not opposition) against TfL sharing data. In the past they’ve been genuinely worried about giving people access to information about their operation. However, I wonder whether this is just a test release of something they’re committed to, but want to do properly.

    Is this definitely data that’s not available anywhere else? Could they be scraping it from the journey planner somehow? It would be very interesting to get an iPhone onto a wireless network and sniff the app’s requests.

    Thanks for the write-up – very interesting!

  3. Jonathan Raper Post author

    Thanks guys for the comments. I will pursue these questions as and when I get the chance with the London DataStore team. Meantime I am testing MyTfL to see what data it really does have access to.


  4. Pingback: TfL withdraw their Journey Planner API… then restore it under pressure « Placr News

  5. Pingback: The case of the disappearing APIs | Transport Innovation

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>